We’re excited to announce our collaboration with Primus, integrating their zkTLS solution with Phala’s TEE framework to push the boundaries of secure, decentralized data interactions.

Building a More Secure zkTLS with TEE

By adopting Phala’s TEE infrastructure, it ensures that no party, including Primus, can tamper with the data verification process. This is a game-changer for security-critical applications such as decentralized finance (DeFi) and AI-driven systems, where trust is non-negotiable.

How It Works

In zkTLS, the attestor acts as a verifier, overseeing encrypted data flows to confirm their authenticity. The challenge? Minimizing trust assumptions on the attestor itself. That’s where TEE technology comes in.

By leveraging Phala’s DStack, Primus attestors now run inside a secure, isolated TEE environment, ensuring that every zkTLS proof is backed by a TEE-issued attestation (quote). This provides end-to-end integrity—even if the OS or application layer is compromised. Plus, anyone can independently verify the zkTLS proof using Phala’s TEE attestation explorer.

And the best part? This hybrid zkTLS solution doesn’t trade off performance for security—TEE acceleration keeps latency low, delivering cloud-like efficiency without centralized trust.

Why Phala?

Phala Network is a leading TEE infrastructure provider, securing AI and blockchain applications with Intel & Nvidia-backed TEE technology. Our Redpill and private-ml-sdk, co-developed with Nearai, enable developers to run AI models like DeepSeek R1 inside TEEs—ensuring secure, private AI computation for Web3 and beyond.

Why Primus?

Primus is a pioneer in cryptographic security, specializing in zkTLS, Fully Homomorphic Encryption (FHE), and the FHE Virtual Machine (FHEVM). Their work on verifiable, encrypted data interactions is shaping the future of privacy-preserving decentralized AI.

Looking Ahead

This collaboration is about more than just zkTLS—it’s about building the foundation for trustless Web3 applications. By merging soft cryptography with hardware security, we’re creating a new standard for decentralized trust—for both humans and AI agents.